Models based on more than one variable of interest form the basis of cokriging. In Brief, Cokriging is a statistical interpolation method that uses data from multiple data types (multiple attributes) to predict (interpolate) values of the primary data type (primary attribute). Cokriging also provides standard errors of the predictions. See also kriging. In Detail, Cokriging is a moderately quick interpolator that can be exact or smoothed depending on the measurement error model. Cokriging uses multiple datasets and is very flexible, allowign you to investigate graphs of cross-correlation and autocorrelation. Cokriging uses statistical models that allow a variety of map outputs including predictions, prediction standard errors, probability, etc. The flexibility of cokriging requires the most decision-making. Cokriging assumes the data come from a stationary stochastic process, and some methods assume normally-distributed data.
